The dauer is an ecologically regulated developmental stage observed in many free-living nematode species and hence provides an interesting model to investigate how ecological adaptations are integrated into developmental pathways during evolution . Extensive studies in C. elegans have uncovered the genetic regulators of dauer formation and comparative studies from other free living nematodes have begun to provide insights into evolution of dauer regulatory genes and pathways [33, 34, 48]. P. pacificus presents an ideal model for comparing dauer larvae of free living nematodes. In this study we provide a comprehensive comparison of gene expression in the dauer stage and dauer exit (12 hour post induction) of the two nematode model systems P. pacificus and C. elegans using the Agilent microarray platform. While we are aware of the available C. elegans expression data on dauer development (e.g. [5, 6, 8–14], we felt that it was necessary to generate C. elegans data de novo for two reasons. First, we wanted to use the same platform as that used for the analysis of P. pacificus in order to enhance the power of a direct comparison. Second, we wanted to benefit from technical advances in custom and long nucleotide microarrays. Based on cross-species comparison of transcriptomes of the ecologically important dauer stage, we draw four major conclusions from our studies.
First, we provide a list of similarities and differences between P. pacificus and C. elegans, which shows an unexpected level of divergence at transcriptome level, even though the dauer stage and recovery process appear to be developmentally conserved. While this comparison allows several evolutionary conclusions (see below), the P. pacificus data set on its own can be used as a starting point for a functional analysis of the dauer stage and dauer exit. This data set represents an invaluable resource given the importance of the dauer stage for the ecology of this nematode. The association of P. pacificus with scarab beetles is restricted to the dauer stage as long as the beetle is alive [27, 28]. Only after the beetle´s death, the nematode exits from the dauer stage to feed on the microbes, which develop on the carcass of the insect. Thus, the P. pacificus dauer stage has a well-defined ecological niche and the expression profiles described in this study will serve as an entry point to future functional studies. For example, our transcriptomic data identifies many P. pacificus specific genes as upregulated in the dauer stage, which implies a potential function in adaptations enabling survival on beetles.
Second, we show that metabolic differences exist between both species, with different patterns of regulation of genes involved in the central carbon metabolism pathways (glycolysis, tricarboxylic acid cycle and oxidative phosphorylation) of P. pacificus upon dauer exit. This difference does not result in obvious phenotypic changes during dauer exit, but may be linked to the differences in life history traits. This is especially relevant given that P. pacificus is adapted for longevity . Under experimental conditions, P. pacificus survives for up to one year in the dauer stage, whereas C. elegans N2 dauer larvae die after approximately 22 weeks. Further studies will reveal how much of the observed metabolic differences are explained by different life-history traits versus differences in rate of development.
Third, this study provides the first nematode evo-devo comparison looking at the downstream consequences of homologous developmental processes between species belonging to different nematode genera. While detailed studies between nematodes as distinct as C. elegans and P. pacificus have investigated the regulation of vulva and gonad development, sex determination and dauer formation [30, 34, 49, 50], most of these studies are concerned with the regulatory mechanisms rather than the “executional programs” of the corresponding developmental processes. The divergence in the expression profiles of C. elegans and P. pacificus adds an important new finding to the growing literature of evo-devo. Previous studies have indicated the limited conservation in the genetic and molecular control of developmental processes in P. pacificus and C. elegans. For example, vulva induction relies on different signaling pathways, requires a novel regulatory linkage and the acquisition of novel protein domains in P. pacificus Wnt signaling . This type of result has been discussed as an example for the theory of developmental systems drift, which proposes that conserved developmental and morphological structures can be regulated by largely diverse regulatory mechanisms . Considering that gene regulatory networks are hierarchically structured, with possibly different rates of evolution at the top level regulatory genes and the most downstream level of effector genes , it can be argued that in principle, developmental systems may diverge due to differences/drift at any of these levels. Unfortunately, unlike the top-level regulatory network, the downstream effector programs of vulva development have largely escaped identification by developmental genetic approaches and have not been easily accessible to transcriptome studies as they are single-cell or small group of cell responses. Our study circumvents this limitation because dauer formation is a “whole body response” of the organism to harsh environmental conditions.
The dauer context is also interesting because in P. pacificus, the key transcription factors of the dauer regulatory network, DAF-16 and DAF-12, are conserved [33, 34]. However, this in itself does not indicate the extent to which the downstream targets of the regulatory network are subject to evolutionary change. Herein, we could demonstrate for the first time that the core downstream execution program of a developmental stage can differ tremendously between P. pacificus and C. elegans, in spite of conservation of upstream regulators like DAF-16 and DAF-12. Thus, these observations make a case for extending the concept of developmental systems drift to the downstream molecular execution of specific developmental stages.
The fourth conclusion is also related to evolutionary theory. While future studies will have to reveal how much of the observed differences between P. pacificus and C. elegans is really of functional importance, it has often been assumed that such differences might simply be neutral [23, 53]. Gene expression, which is neither strongly deleterious nor advantageous, previously termed “gratuitous expression" , will not be under selection and will be free to evolve by drift. Consequently, such expression is probably not functional. However, such arguments may not apply to the dauer stage since nematodes live off their internal limited energy resources and any random or neutral transcriptional activity would diminish these limited resources.
Comparative studies in developmental genetics have driven the studies on evolution of developmental mechanisms, and with whole genome sequencing of many animal species has now highlighted new facets of evolutionary dynamics through comparative genomic studies . Since transcriptional regulation is a key building block in the genotype to phenotype translation, comparative transcriptomic studies add another dimension to the analysis of evolutionary processes . Our work contributes to the growing set of results from comparative transcriptomics in diverse developmental systems (e.g. [54–58]). These studies together span a range of conclusions, from high transcriptomic conservation at one end, to relatively low conservation in others, suggesting inherent constraints as well as flexibility in the evolution of gene regulatory networks . Future studies comparing transcriptomes of homologous biological processes in related species, will be important for understanding the role of transcriptome evolution in generating animal diversity. Ultimately, this will also reveal the extent to which the conserved or divergent expression changes are subject to adaptive and non-adaptive forces during evolution.